The relationship between parental smoking and lower respiratory tract illness (LRI) was studied in a large cohort of infants followed prospectively from birth. Illnesses were diagnosed by physicians using agreed-on criteria, and parental smoking histories were obtained by questionnaire. The LRIs were differentiated into wheezing and nonwheezing episodes, and the age at first illness of either type was evaluated in relation to smoking by parents. The odds of having an LRI were significantly higher in children whose mothers smoked (odds ratio 1.52; confidence interval 1.07 to 2.15). The odds were higher if the mother smoked a pack of cigarettes or more per day and if the child stayed home rather than attending day care (odds ratio 2.8; confidence interval 1.43 to 5.5). Logistic regression indicated that the LRI rate was significantly elevated both in children exposed to heavy maternal smoke in the absence of day care, and in those who use day care but were not exposed to maternal smoking of a pack or more per day. These findings could not be attributed to other confounding variables. Neither paternal smoking nor smoking by other household members was consistently related to the LRI rate. The relationship of maternal smoking to LRI rate was evident for both wheezing and nonwheezing illnesses. Maternal smoking of a pack or more per day was also related to an early age at first LRI, for both wheezing (p less than 0.05) and nonwheezing (p less than 0.002) illnesses. In sum, maternal smoking is associated with a higher rate of LRIs in the first year, particularly when mothers smoked a pack or more per day and when the child did not use day care.